Coal Science - American Chemical Societyhttps://pubs.acs.org/doi/pdf/10.1021/ba-1966-0055.ch027Similarnumerical values o...
2 downloads
89 Views
1MB Size
27 Kinetics and Mechanism of Solution of High Volatile Coal
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
GEORGE R. HILL, HASSAN HARIRI, R. I. REED, and LARRY L. ANDERSON Department
of Fuels Engineering,
University of Utah, Salt Lake City,
Utah
A kinetic study of the dissolution of a Utah high volatile bituminous coal in tetralin has been conducted. Equipment for obtaining rate data during the early stages of the reaction has been developed. The data are interpreted in terms of a pseudo second-order rate constant, average heats of activation, and apparent entropies of activation, the numerical values of these functions depending on the degree of extraction. The rate of solution in tetralin of this coal is not an equilibrium phenomenon but is a kinetically controlled reaction in which the average activation energy increases as the reaction proceeds.
y y h i l e solvent extraction is used to convert coal to valuable l i q u i d a n d gaseous products, it also may be used to study coal structure a n d composition. Thermal dissolution of coal (solvent extraction w i t h a reaction temperature above the boiling point of the solvent) with various solvents has been studied extensively for a number of years (5, 9, 10, 12, 13). In fact, a voluminous literature has been presented and reviewed which deals directly w i t h the extraction of coal and extraction conditions. This paper describes experimental result of kinetic studies together with a new approach to the theory a n d mechanism of solvent extraction. Experimental
Procedure
T h e coal used for this study was taken from a working face of U t a h Spring C a n y o n C o a l M i n e . T h e analysis as given by Commercial Testing a n d Engineering C o . for the sample (calculated on a dry basis) is shown below. T h e coal was ground i n the laboratory a n d sized to pass 200 mesh. A s h determinations on the ground coal sample were made, a n d an average value of 5 . 2 % was used to correct for the ash content of the extracted coals. 427
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
COAL SCIENCE
428 Ultimate Analysis Carbon Hydrogen Nitrogen Oxygen Sulfur Chlorine
Proximate Analysis
72.88 5.58 1.51 10.82 0.65 0.19
Water Ash Volatile Matter Fixed Carbon
— 8.37 45.71 45.92
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
Heating value—13,237 B.t.u./lb.
Ol
I
.05
0.1
I
0.15
I
I
I
0.2
0.25
0.3
C/S Cool, grrt/Solvent, ml. Figure
I.
Determination optimum
of coal/solvent extraction
ratio
for
T o find the optimum ratio of coal/solvent ( C / S ) , a series of solubility experiments were performed at constant temperature and time but w i t h differ ent C / S . It was found that at ratios smaller than 1:8 the quantity of coal dissolved d i d not increase (Figure 1). A l l kinetic studies were made at ratios of 1:10 to ensure that excess solvent was always present. F o r each r u n , coal samples of approximately 50 grams were dried, at 1 0 0 ° C . for 4 hours and weighed after cooling i n a desiccator for % hour. In the early experiments, coal and solvent were mixed i n the autoclave, and runs were performed. It was found that the time necessary for the autoclave and mixture to be heated from room temperature to reaction temperature was 1 % - 2 hours. W h e n extraction fraction vs. time was plotted, it showed that at higher temperatures more than 8 0 % of the total possible extraction of coal dissolved before the system reached the reaction temperature. Consequently, the data obtained i n the first 2 hours were incorrect. T o overcome this problem a coal injector was designed and constructed. T h e coal injector was a stainless steel cylinder, 8 inches long, 3 inches outside diameter, and 1.5 inches inside diameter. One side of the cylinder was open, and the other was provided w i t h a 0.25-inch stainless steel female fitting (Figures 2 a n d 3 ) . A piston, provided w i t h an ' Ό " ring gasket, was used for die open side of the cylinder and could move back and forth b y a mechanical
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
27.
H I U ET A l .
Kinotit* of Solution
14
5'V
l ^ i e — * — ι ιβ-
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
5 /
429
V
7 "
—Jf
—-·,
Τ
— L -
τ 1
ï
-M THREAD CAP 3/165.
PRESSURE
P A D —
"0*RING RETAINER SQUARE
REFLUX'O" RING
,
3 7"*·
18 N . R T .
CYLINDER Figure 2.
Major parts of coal injector with their
il
measurements
screw system. T h e outside body of the coal injector's cylinder was graduated and calibrated in cubic centimeters injected per linear inch drive of the cylinder. T o use the coal injector, equal amounts of coal a n d tetralin (approxi mately 50 grams of coal which was prepared by the above mentioned pro cedure a n d 50 cc. of tetralin ) were used i n each run a n d mixed i n the coal injector. W h i l e preparing coal paste i n the coal injector, 450 cc. of tetralin were being heated i n the autoclave to reaction temperature. W h e n the auto clave containing tetralin reached the reaction temperature, the coal paste was injected into the autoclave through a fitting on the autoclave. T h e temperature was carefully observed during this operation, a n d it was found that after 1 or 2 minutes the system returned to the reaction temperature ( F i g u r e 4 ) .
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
COAL SCIENCE
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
430
Figure 3.
Coal injector used for solvent
extraction
The autoclave used for this study was a 1-liter capacity Magne D r i v e auto clave made by Autoclave Engineers. It is provided w i t h a l i q u i d sample line, a gas sample line, quenching tubes, Magne Drive stirrer, pressure gauge, heat ing jacket, and two extra connections for special uses. One of these connections was used for injecting the coal paste into the autoclave. T h e autoclave design pressure of 5000 p.s.i.g. p r o v e d adequate for all runs. E a c h run was continued for about 2 5 - 3 0 hours, and about 25 liquid samples were taken during this time. T h e volume of each sample was between 15 and 30 cc. Samples taken i n each run were treated similarly. T h e y were transferred into Soxhlet thimbles which had already been dried for Vz hour. T h e thimbles were placed i n the Soxhlet extraction unit using benzene as the solvent. T h e extraction was continued until the l i q u i d circulating i n the unit became clear and colorless (indicating pure benzene). T h e thimbles were then removed from Soxhlet extraction tubes and placed i n the oven at 1 0 0 C . for % hour and weighed. T h e difference between the weight of the thimbles before and after Soxhlet extraction was the weight of residues. F o r each r u n , the ratio of coal i n grams to the mixture (coal plus tetralin) i n cubic centimeters used i n the experiment was carefully measured; to obtain the initial amount of coal i n each sample taken, that ratio was multiplied by e
ο
δ
50
ÏOO
150
200
250
300
350
400
Time in Minutes Figure 4.
Variation
of temperature during process of solvent at 440*C.
extraction
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
27.
HILL ET AL.
Klnotics of Solution
431
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
the volume of the sample. This procedure is valid if the following assumptions are accepted: (1) T h e mixture in the autoclave is homogeneous. This was insured by keeping the Magne D r i v e stirrer always at a steady 1500 r.p.m. It was experimentally proved that even distribution of coal and tetralin i n the autoclave is a fair assumption because the fraction extracted from the last portion of mixture, w h i c h remained i n the autoclave and was taken after the system cooled to room temperature and the autoclave was opened, was found to be very close to the yield obtained from the last sample taken from the system through the sample lines at reaction temperature. (2) The volume of the coal before and after extraction does not change very m u c h . This also was found to be a fair assumption because i n a l l the runs the initial and final volumes were accurately measured and gave C/S ratios of approximately 1/10. Generally, the total volume after extraction was about 20 cc. less than the initial volume of the mixture. Since some of the coal converts to gas and some adheres to the wall of the autoclave, it is concluded that the density of the mixture at the start and end of the run d i d not change significantly. However, even the ratio of coal (grams)/total volume ( c c . ) , ( C / V ) , for this small change was corrected.
0.1 0 Figure 5.
Theory and
200 Time-yield
400 600 800 t,Time in Minutes curve for thermal dissolution in tetralin
1000 of bituminous
1200 coal
Mechanism
T h e conventional mechanism and mathematical treatment for solvent extraction kinetics was proposed by Oele et al in 1951 (11) and has since been accepted and used by others. Oele assumed that the coal w i l l enter the l i q u i d solvent in accordance with a zero-order rate law u p to a certain time;
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
432
COAL SCIENCE
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
then not only is the zero-order reaction maintained, but now a reverse reaction occurs, which was assumed to be unimolecular. W h i l e interpreting our experimental data (Figures 5, 6, 7) by this and other mathematical expressions, we concluded that there are other mechanisms w h i c h represent the data better than O e l e s model. T h e tetralin solutions were found not to be saturated with dissolved coal i n the experiments performed by us. W h e n In ^ 1 —
^
was plotted vs. t, a straight line was obtained by
the method of least squares; it d i d not pass the second boundary condition of O e l e s equation—i.e., at time zero, χ was not zero. Chariot ( I ) used this model, and his graph of In ^ 1 —
χ" )
v s
' **
m e
<
*
o e s
n o
*
s e c o n <
^
boundary condition of Oele ( I I ) . T h e dissolution of coal by solvent action raises several problems as to the probable mode of action of the solvent and the possible kinetic consequences of such reaction.
0
200
400 600 800 t Time in Minutes
1000
1200
y
Figure
6.
Time-yield
curve for thermal dissolution tetralin
of bituminous
coal in
A coal particle can be considered as made up of a main structure per meated by both macro and micro pores and other materials which are lodged i n the coal and are accessible to the action of the solvent, tetralin. This con ventional picture of coal has received some experimental support (3, 8). Moreover, good evidence exists that the pores present range i n size from an ultrafine structure having gaps a few Angstroms wide (the result of the ran d o m packing of relatively large channels) u p to 100 A . w i d e (14). Still larger capillaries and fissures are present. Dissolution of such coal can occur
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
27.
HILL ET AL.
433
Kinotks of Solution
200
400 600 800 t Time in Minutes
1000
1200
9
Figure
7.
Time-yield
curves for thermal dissolution
of coal in
tetralin
in several ways, and some of them have distinct kinetic consequences. These include: Kinetic Rate Consequence Constant (1) Dissolving out of included materials (2) Dissolution of the coal structure in the presence of a large volume of solvent (3) Diffusion out of the micropores (4) Hydrogen transfer reactions (5) Solvent imbibition
1st and/or 2nd 1st
k. k
2nd 2nd 1st or 2nd
k, k,
f
kf, k.
Dissolving out the organic materials which are not part of the main coal structure depends upon the size of the micropore. It may become the ratecontrolling step and represent the diffusion of the solution of the organic material i n the solvent tetralin, from out of the pore, which implies that the process of solution w i l l be first order w i t h respect both to the coal a n d solvent. If, however, the material dissolves i n a large port or fissure, then the diffusion process may be so rapid that the reaction w i l l be first order w i t h respect to the coal only. T h e hydrogen transfer reaction w i l l be a second-order process since the transfer reaction from the tetralin to the coal w i l l not be an abundant process at the temperatures used here. Solvent imbibition w o u l d result i n swelling of the coal particles w i t h its consequent effect upon the porous structure. This consideration does not seem important i n the present experiments since little swelling of the coal has been reported (4). It is known that solvents w h i c h swell the matrix can release
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
434
COAL SCIENCE
the materia] (a colloidal matrix of bonded or interlocked units) which forms the residue of the low temperature extraction. At higher temperatures the nature of the solvent is much less critical, although peptization can sometimes occur. Since more than one of these dissolution processes might occur in the coal extraction experiments, it is necessary to allow for concurrent chemical reactions when constructing a rate equation. Since reactions are either first or second order, a kinetic expression having concurrent reactions of first and second order must be derived. +
j
t
=kt(a — x)+k.(a
— x){h — x)
Here, a represents the initial concentration of coal, b that of tetralin; kt is the first-order reaction rate constant, and k, is the second-order constant. The other symbols have their usual significance. Rearranging: kt(a-x)
+ dx +k(a — x)(b — x) = + dt T
The quadratic denominator has the roots: x
r=
h + k.b , —ζ and a. k.
Integrating by parts after substitution and between the limits χ = 0 , t = 0 ; χ = x, and t = f, one obtains:
When t ~ 0 , χ ~ 0 , and therefore in the initial stages of reaction the expression simplifies to: In
~ (*,_lfc.fl+U>) t
When the reaction is near completion, χ — a. Then log
g
g
χ
changes
rapidly with small variations in x, and one may consider the bimolecular con tribution as overshadowed by the first-order reaction. This follows since fc, the concentration of the solvent, which is in excess, is much greater than a, the concentration of coal. ( v
E
)
'
In ( * ' t ^ - * · * ) ( -^—) ~ \ kt + k.b ) \a — χ)
j - ^ r + In — ° kr + k.b a— χ
The complete rate equation (E) reduces to the following approximation on substituting, simplifying, and returning to Briggsian logarithms: 2.303 l o g
=
J
* ±
T
t
+ t(kr -
k.a +
k.b)
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
27.
HILL £7 AL.
Ktnotk* of Solution
435
If k.b ~ kt, the equation simplifies to: 2.303 l o g
=
±
+ t(k.b — k.a + kt)
F r o m this research, kt = 2.26 X 10"* sec." data kt = W*'*e * . M,
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
coal, k. is 9 . 1 0 V
2,,/
1
F r o m an Arrhenius plot of the
F o r the extraction of the interspersed materials i n the
T
*. r
A detailed analysis of the data i n terms of the simplified equation ( E ) demonstrates that the constant kt for the "chemical" dissolution is an average value a n d that the process involves a series of reactions w i t h increasing activation energies. Considering the fundamental structure of coal as proposed b y H i l l a n d L y o n ( 7 ) , the high volatile coals consisted of large alkylated, polynuclear, oxygenated, aromatic, a n d heterocyclic nuclei. These are held together b y oxygen a n d sulfur atoms cross-bonded to carbon-carbon bridges and threedimensional tetrahedral bonds w h i c h form the basic porous structure of the coal matrix. W i t h this structure i n m i n d it is believed that at the beginning of reaction the materials which enter the l i q u i d phase are those w h i c h were trapped i n the coal pores a n d w h i c h may be weakly bonded to the main coal structure. These require the lowest activation energy to separate from the matrix of the coal. T h e remainder of the coal dissolves with the breaking of chemical bonds, consequently requiring a greater activation energy, a n d as the extraction continues the activation energy of extraction increases. This is illustrated b y the following: *i C o a l -> R i + L i + G i
(1)
R i -> R + U + G2
(2)
R2 -> Rs + U + Gs
(3)
*. R . - » R*+i + L . + i + G-+i
(4)
2
where k\ > fa > fa > . . . > fc. In the first step, a l l the coal bulk which is readily available i n the mixture of coal a n d solvent goes to R i which is solid, L i w h i c h is l i q u i d , a n d G i w h i c h is gas. T h e rate constant for this unimolecular reaction is ki. W h e n Reaction 1 is well advanced, Reaction 2 becomes the main route w i t h the extraction of R i (rate constant )b) a n d so o n . According to D r y d e n ( 2 ) , extraction occurs b y the removal of units of colloidal size directly from the coal, a n d he proposes a model for extraction as follows: A matrix of larger a n d more strongly linked micelles, w h i c h can be partly and progressively dispersed i n a suitable solvent b y increasi n g the temperature, is intimately associated with a proportion of
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
436
COAL SCIENCE
smaller, less strongly bonded micelles that are normally trapped within this matrix unless it is first swollen by the solvent. The difference between these two classes of micelles was supposed to be one of degree rather than k i n d . Possibly they formed a continuous series, the dividing line being determined by the temperature. The micelles, assumed to be rigid and comparatively indestructible, w o u l d be extracted as individuals by diffusion through the swollen pores of the matrix. Dryden's model may describe the extraction at l o w temperatures, but at temperatures above 3 5 0 ° C . the assumption of indestructible micelles is not tenable. Apart from the idea of an indestructible micelle, Dryden's model is similar to the one described above. Moreover, i n this study, no distinction is made between the colloidal part of the coal and the lighter hydrocarbons. O f course it was observed that the liquified coal extracted at 3 5 0 ° C . had a higher average molecular weight than the one at 4 5 0 ° C . Distilling these samples showed that extracts taken at lower reaction temperatures have higher boiling points. In the above proposed scheme it is assumed that at any instant one of these steps predominates and all the reactions are forward reactions only. If there are any reverse reactions, they must have very slow rates and can be neglected. T o support the above assumptions—i.e., that the reverse reaction is negligible and that there is no equilibrium between forward and backward reaction—the following experiment was performed. A sample of 50 grams of coal and 500 cc. of tetralin was placed i n the autoclave, and the temperature was raised to 4 4 3 ° C . After 3 nours at this temperature, a l i q u i d sample was taken. T h e temperature was then lowered to 3 5 0 ° C , maintained for about 6 % hours, and another sample taken. This process was repeated twice w i t h the following results (Table I ) : Table I.
Experimental Results Testing Reverse Order Mechanism in Solvent Extraction of Coal 8
Sample No. 1 2 3 4
Wi 3.6457 2.7296 4.5217 4.3855
w
W 3.2007 2.4401 3.9877 3.8437
f
0.4450 0.2895 0.5340 0.5418
AW W — W 0.93 3.4561 2.5877 0.94 4.2866 0.93 4.1574 0.926 c
x
r
t 180 500 850 1450
T, °C. 443 350 267 246
• A l l samples heated initially to 433 C. and kept at that temperature for 180 minutes. e
Table I shows that the fraction extracted is constant within experimental error. Hence, one may conclude that there is very little, if any, reverse reaction involved. A similar procedure was employed in other experiments, and the same result was obtained. T h e model developed here assumes that only unimolecular reactions are involved, but that the "first-order reaction velocity constant" varies w i t h the fraction extracted. T o derive a suitable mathematical relationship between the first-order reaction velocity constant, k, and the fraction extracted, x, one proceeds as follows: dx
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
27.
HILL E7 AL.
437
Klnotlc* of Solution
Froction Ettrocted Figure 8.
Fraction
extracted vs. the rate constant
dx/dt . ι = k 1 —χ A t each temperature, the method of finite differences was used and -· evaluated for a series of x. T h e plot of (Δχ/Δ*)/1 — χ = k vs. χ closely approxi mated a straight line (Figure 8 ) . It is therefore considered that the rate constant, k, changes linearly with x, the fraction extracted. k = Ci— * = C i ( l — If then,
C x 2
g x )
C i = ko and C2/C1 = a k = Jfc#(l —
ax)
T h e parameters Jfc. and a were found experimentally (Table I I ) ; a is found to be the reciprocal of the maximum possible extraction, (xm), at a given temperature, ( t ) ; &· was the intercept.
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
438
COAL SCIENCE T h e above equation then becomes: — o x ) ( l — x)
%- = k.(l
where k* is a pseudo second-order rate constant. ( l _
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
h l
Table II.
a
x
f ( l _
(r=^)
=
x
) = ^
*
(
+ C
™
d
, i + c ) ( e _ 1 )
Comparison of fc, and a in k = fc, (1 — ax) in Integrated Form 1 Xmmx
T°C.
ko
a
350 375 400 410 420 430 440 450
0.0059 0.0120 0.0372 0.0542 0.0760 0.110 0.142 0.2210
1.390 1.142 1.111 1.092 1.088 1.067 1.064 1.057
1.387 1.188 1.111 1.099 1.088 1.066 1.061 1.042
U s i n g the boundary conditions: t = 0, χ = 0, a n d t = oo, ax = 1.0; substituting the first boundary condition C = 0, the final equation becomes
' " ( T ^ T X ) ^ * -
1
)
F o r a l l temperatures from 350° to 450°C. Jfc. was calculated from the above equation for a series of χ a n d t a n d then averaged. T h e k. average was fairly close to the intercept of k vs. x. F r o m the original χ a n d t, the curve w h i c h best represented the experi mental data was plotted, a n d the data used for calculating the k»'s were taken from that curve. T h e maximum rate constant is represented b y k.. A n Arrhenius energy of activation, £ , a n d E y r i n g enthalpy of activation, ΔΗ*, were obtained by plotting In h and In (ko/T) vs. l/T. T h e values were 32.0 kcal./mole for the Arrhenius activation energy and 31.0 kcal./mole for the E y r i n g activation enthalpy (Table III and Figures 9 a n d 10). Table III. T, °K. 623 648 673 683 693 703 713 723 • Eo =
l/T 16.050 15.428 14.859 14.640 14.430 14.225 14.025 13.831
Χ 10' Χ 10' X 10Χ 10" Χ ΙΟ" X 10" X 10" Χ ΙΟ" 4
4
4
4
4
4
4
31.8 kcal. for Arrhenius plot;
Final Results of Solvent (k./T)
ko
8.988 Χ ΙΟ"* 17.1162 X 10-« 51.0407 X 10"* 79.6709 X 10"° 107.5035 X 10^ 153.8437 X 10"* 162.9705 X 10-* 275.375 X 10"·
0.0059 0.0120 0.0372 0.0542 0.0760 0.1100 0.1420 0.2210
&Hot =
30.8 kcal. for Eyring plot
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
27.
HILL ÎT AL.
Kinotln of Solution
499
-β
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
10
12
Ο Integrated Form k = r - r — In 0
-14
• Difference Form
A
,
—
A A
"0 -16
138
142
146
1
150 (l/T)
Figure 9.
Plot of k./T
154
ΙΟ
for calculating
158
162
3
the enthalpy of
activation
T h e reaction rate constant at the beginning of the experiment—i.e., at χ = 0—is k»; it should not be mistaken for a zero-order rate constant. T o investigate the quantitative variation of enthalpy of activation a n d entropy of activation w i t h x, a series of rate constants w i t h different tempera tures for each value of χ were taken, and b y using the theory of absolute reaction rates (6) : h where k* is the Boltzmann constant a n d h is the Planck constant. Values for the average ΔΗ* a n d the apparent AS* were obtained for four different values of χ a n d are given i n Table I V . A t the beginning of the reac tion the apparent entropy of activation was negative, but as the reaction con tinued, it increased to zero and then became positive. A n explanation for this Extraction of Coal with Tetralin" (k./T) 9.4695 18.5040 57.2855 78.0 109.8680 156.4750 199.155 305.6651
Χ ΙΟ"* X 10X 10X 10X 10X 10X 10X 10-
ln(k.)
ln(k./T)
--5.2728 -4.4232 —3.2941 —2.9150 —2.5770 —2.2072 —1.9519 —1.5559
—11.5674 —10.8923 —9.8037 —9.4416 —9.1155 —8.7633 —8.5200 —8.1400
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
C O M SCIENCE η
1
1
1
J
1
«
I- χ \) ]—5χ Ln
0
ο Difference Form k =
r
1
I Ο Integrated Form k = ^ 1
Δχ / At
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
0
146
150 (Ι/Τ)Ι0
Figure 10. Arrhenius
154 3
plot for calcufoting
the activation
energy
behavior based o n an increase i n number of sites available for reaction is being checked a n d w i l l be published. Solvent extraction of coal can be explained b y a series of independent first-order reactions, the residue of one reaction being the reactant for another reaction. It can be formalized as follows: ko C o a l ?± Xo* -> R i + L i + G i Jfci' R i ?± X i * -> R + L + G 2
2
2
fc ' R ?± X * -> R + L a + Ga kn 2
2
R» ^± X**
2
3
R«+i + L«+i + Gn+i
W h e r e X*'s are activated complexes for the reactions. These reactions should not be confused w i t h consecutive reactions since i n this case R i , R , and R* are the remaining unreacted coal rather than a new product. 2
Table IV.
Γ, °K. fci/Ti *2/T
2
k /T< 4
ke/Te k /T k /Ts 7
H
7
623 648 673 683 693 703 713 723
Variation of Enthalpy and
1/T(°K.)-' 1.605 1.542 1.486 1.464 1.443 1.422 1.4025 1.383
1010101010101010-
χ= 0 0.0947 0.1850 0.5527 0.7935 1.0907 1.5647 1.9915 3.0570
ΔΗί (kcaL/mole) AS* (cal./mole °K.) apparent
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
X X X X X X X X
io-< io-< ίοίο10 10 1010-
37.2 —19.5
1
4
H
H
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
27.
HILL ET A l .
Figure
441
Kinetic* of Solution
11.
Flot of k/T vs. l/T for activation enthalpies
evaluating
The value of — 1 9 . 5 e.u. for the apparent entropy of activation obtained at the beginning of the extraction includes a term for the limited number of sites where dissolution could occur (Table I V a n d Figure 1 1 ) . It is considered probable that the dissolution of material from the pores (process Rr) occurring simultaneously can account for part of the large negative value. Chariot ( I ) Entropy of Activation with Fraction Extracted χ = 0.5 0.0290 0.07514 0.2459 0.3579 0.5003 0.7312 0.9327 1.4661
X Χ Χ Χ Χ Χ Χ Χ
40.0 —16.8
χ = 0.8
10" ΙΟ" ΙΟ" ΙΟ" ΙΟ" 10" 10" 10"
χ = 0.9
4
4
4
4
4
4
4
4
0.0091 0.0618 0.0966 0.1426 0.2310 0.2973 0.51170
Χ Χ Χ X Χ Χ Χ
51.8 —2.1
10" 10" 10" 10' 10" 10" 10"
4 4 4
4 4 4
4
0.0095 0.0234 0.0640 0.08555 0.1936
Χ Χ Χ Χ X
10" 10" 10" 10" 10" 4
4
4
4
85.5 41.0
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
442
COAL SCIENCE
found that a two-stage mechanism fits the lower temperature range of dissolution well. To determine whether the low temperature "physical" dissolution could be separated from the processes with higher activation energies, the following experiments were conducted. Coal and tetralin with a ratio of about 1 gram of coal to 10 cc. solvent were mixed in the autoclave, and the temperature was raised to 2 5 0 ° C . This was the temperature at which transition of a diffusion process to a surface reaction process takes place as postulated by Chariot (I). The contents of the autoclave were divided into three parts and placed in three thimbles, which were already treated and weighed according to the procedure previously described (Figures 12 and 13).
100
200
900
400
500
600
t,Time in Minutes Figure 12. Variation of fraction extraction with time for the second stage of two-stage solvent extraction of coal; first stage at 250°C. for 8 hours not shown The thimbles were treated in the Soxhlet extraction apparatus using benzene. The residue left in the thimbles was dried and carefully weighed. A total of 12% coal was extracted in this manner. These three coal samples were extracted with fresh tetralin at reaction temperatures of 4 0 0 ° , 4 2 0 ° , and 440 C. The experimental procedure was the same as previously described—i.e. part of the tetralin was heated in the autoclave to trie selected reaction terne
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
27.
HILL ET AL.
Kinetics of Solution
100 200 300
443
400 500 600
700
t Time in Minutes 9
Ο One stage at T= 440 C. #
• Two stages (first at 250*0, second:440 C) #
Ο Chariot s data for 250 C #
Figure 13.
Table V. T, °C. 440 420 400
Comparison
extraction
Final Results for Second Stage of Two-Stage Solvent Extraction of Coal" l/Xn,
0.84 0.80 0.79
of one-stage and two-stage of coal
( « - £ )
1.19 1.25 1.265
ΔΗ* in kcal. AS* (for * = 1.0)
1.1914 1.276 1.266
k' for χ = 0
k* for χ = 0.7
l/T
0.1914 0.0506 0.0140
0.0362 0.0060 0.0014
0.0014025 0.0014430 0.0014a59
60.0 -1-6.2 e.u.
75.0 4-25.5 e.u.
• First stage—250*C. for 8 hours; 12% was extracted and is not shown here..
perature a n d the coal paste which was prepared by mixing the coal sample with the remaining solvent i n the coal injector was then injected into the reactor. A series of samples was taken to determine the amount of residue left. The data obtained were treated exactly as those obtained i n former experi ments, a n d a series of average enthalpies of activation a n d apparent entropies of activation was obtained for different x's (Table V , Figures 14 a n d 15). It was found that removing only 1 2 % of coal during the first stage of extraction increased the initial apparent entropy of activation from a negative to a positive value. It further shows that this value increases as χ increases. This is expected if the surface area available for reaction increases during the course of the reaction.
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
444
COAL SCIENCE
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
T h e experimental results confirm the idea that at the initial stage of the experiment, the reaction is under diffusion control. This process has a very l o w activation energy, and when combined w i t h thermal disintegration of the coal at higher temperatures, it lowers the average free energy of activation for the reaction.
05
0.6 0.7 OA Fraction Extracted, x « - J J —
Figure 14. Variation of rate constant with χ for the second stage of two-stage extraction of coal; first stage at 250°C. for 8 hours not shown Conclusion C o a l is a complicated material, and more than one simple chemical process takes place during solvent extraction. The activation energy necessary for dissolving coal increases with the extent of the process up to a point; at a certain temperature insufficient energy is available for more extraction. W h e n the activation energy necessary for extraction becomes more than maximum energy supplied, additional coal w i l l not dissolve. The rate constants, average heats of activation, and apparent activation entropies predict the dissolution
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
27. HILL ET AL.
4 3 2
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
T c C
445
Klnotics of Solution
i
0 1
08 ftfî
Cl Λ
M
i ο
.02
*
A .001 ° .008 f .006
y .
* .004 .002 .001
1.4
1.45 (I/T)I03 inCK)*
1.5 1
Figure 15. Arrhenius plot for the second stage of a two-stage extraction of coal for two values of x; first stage at 250° not shown rate of this coal i n tetralin. T h e heat of activation increases, and the apparent entropy of activation becomes increasingly negative during the course of the reaction. Acknowledgment Appreciation is expressed to Yacob Shifai, Norbert Kertamus, a n d L a r r y Chariot for their contribution to this paper. T h e research reported here is supported by the Office of C o a l Research, Department of the Interior under Contract N o . 14-01-0001-271 and by the University of U t a h Research F u n d . Literature Cited (1) Charlot, L. Α., Master's thesis, University of Utah, 1963. (2) Dryden, I. G. C., "Chemistry of Coal Utilization," H . H. Lowry, Ed., p. 248, Wiley and Sons, New York, 1963. (3) Dryden, I. G. C., Fuel 37, 444 (1958). (4) D'yakova, M . K., Davtyan, Ν. Α., Bull. Acad. Sci. URSS, Classe Sci. Tech., 1945, 203; J. Appl. Chem. 21, 113 (1948).
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
446
COAL SCIENCE
(5) D'yakova, M . K., "Production of Synthetic Liquid Fuels and Chemical Products by Thermal Dissolution of Solid Fuels," p. 86, Academy of Science of the U.S.S.R., 1957. (6) Glasstone, S., Laidler, K. J., Eyring, H., "The Theory of Rate Processes," McGraw-Hill Book Co., 1941. (7) Hill, G. R., Lyon, L. B., Intl. Eng. Chem. 54, 36 (1962). (8) Huck, Kartweil, Brennstoff-Chem. 34, 97, 129 (1953). (9) Kiebler, M . W., "The Chemistry of Coal Utilization," H . H . Lowry, Ed., Vol. I, p. 724, Wiley and Sons, New York, 1945 (10) Lowry, H . H., Rose, H. J., Bur. Mines Inform. Circ. 7420 (1947). (11) Oele, A. P., Waterman, H. I., Goedkoop, M. L., Van Krevelen, D. W., Fuel 30, 169 (1951). (12) Pilipetz, M. G., Kuhn, E., Friedman, S., Storch, H. H., U.S. Bur. Mines, Rept. Invest. 4546 (1949). (13) Storch, H. H., Chem. Rev. 29, 483 (1941). (14) Van Krevelen, D. W., "Coal," p. 178, American Elsevier Publishing Co., New York, 1961. RECEIVED October 5, 1964.
Discussion George K a p o : In your opinion what is the extent of the available internal area for solution? George R . H i l l : In the l o w temperature " p h y s i c a l " solution process the surface area would probably be that determined by B E T adsorption measurements. In the high temperature process, apparently the coal structure is opened u p , and the surface would be the total surface of a l l the "molecular" units. This occurs, as the dissolution proceeds, by a combination of chemical bond breaking a n d solvent action with hydrogen transfer to the free radicals produced. D r . K a p o : H o w does diffusion influence the kinetics of solution? D r . H i l l : In the l o w temperature solution process the activation energy value suggests that a physical process—probably diffusion—is rate controlling. T h e large ( a n d increasing) value of the heat of activation for the major portion of the dissolution reaction requires that the rate is a chemically controlled process—very likely the breaking of chemical bonds. K u l a i A . K i n i : Y o u said that there was no swelling when coal was extracted with aromatic solvents. W h a t method was used to measure swelling? D r . H i l l : N o direct measurements other than usual observation were made on the swelling of the coal. T h e bulk volume remained unchanged. Norbert Berkowitz: I think the kinetic treatment of the experimental data is of questionable validity. T h e extraction process is evidently accompanied by considerable changes i n the geometry of the coal particles (e.g., swelling a n d dispersion); there is the unresolved question of whether the extract forms a solution or dispersion; finally, there is an obvious but somewhat indefinite effect of coal decomposition. T h e latter point alone w o u l d make determining a temperature effect (and hence, calculating an "activation energy") very difficult practically, if not impossible.
Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
Downloaded by UNIV OF CALIFORNIA SAN DIEGO on January 31, 2017 | http://pubs.acs.org Publication Date: January 1, 1966 | doi: 10.1021/ba-1966-0055.ch027
27.
Hill ET A l .
Klnotlc* of Solution
447
D r . H i l l : T h e point is well made that the solution of coal i n solvents is a complicated process involving separation of macérais, breaking of relatively weak hydrogen bonds, a n d , increasingly at high temperatures, the rupture of covalent chemical bonds. T h e "activation energy" calculated from a plot of log of rate vs. l/T obviously must be an average value for a l l the processes which are occurring. The slope of the line obtained from this plot is, however, independent of the final state of the coal in solution and is meaningful in terms of the " m i n i m u m " average chemical bond strength of those bonds broken i n the dissolution reaction. Observing a low activation energy at low temperatures requires the conclusion that a primarily physical separation process is occurring below 2 8 0 ° C , but that above that temperature the breaking of chemical bonds of increasing strength becomes rate controlling. T h e data also require a conclusion that the degree of solution (and pyrolytic decomposition) is determined by a rate process and is not an equilibrium phenomenon. T h e very high (80 kcal.) average heat of activation obtained at high temperatures suggests that many carbon-carbon bonds are being severed i n the high temperature range. It is agreed that a physical interpretation of the entropy of activation is most tenuous; nevertheless, it is a useful beginning for understanding what is occurring. T h e absolute value of has no meaning unless the initial state is well defined a n d constant. However, the change i n AS* as the reaction proceeds is significant and requires an explanation like the one proposed. If a better method of utilizing the data can be suggested, we should be pleased to apply it. Leslie Reggel: C o u l d you say more about the structure of the coal "molecule" you showed us? Is enough hydroaromatic hydrogen included? D r . H i l l : T h e coal "molecule" (reproduced from Ind. Eng. Chem. 54, 36 ( 1 9 6 2 ) ) does i n fact include hydroaromatic hydrogens. A l l of the R° Ν alicyclic rings (some six or seven i n the diagram) have hydroaromatic hydro gens as does the alicyclic ring i n the lower left hand corner of the model. I n this "average structural unit" some 14.7% of the hydrogen is aromatic, 7 7 % of the hydrogen is alicyclic a n d aliphatic, a n d 8 . 3 % of the hydrogen is i n functional groups. T h e percentage of hydroaromatic hydrogen corresponds to that reported by Given (Fuel 39, 147 (1960) ) .
American Chemical Society Library 1155 16th St., N.W. Washington, OX. 20036 Given; Coal Science Advances in Chemistry; American Chemical Society: Washington, DC, 1966.
